Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network

Rashika Mishra; Ovidiu Daescu; Patrick J Leavey; Dinesh Rakheja; Anita L Sengupta

doi:10.1007/978-3-319-59575-7_2

Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network

Rashika Mishra, Ovidiu Daescu, Patrick J Leavey, Dinesh Rakheja, Anita L Sengupta

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

24 Scopus citations

Abstract

Pathologists often deal with high complexity and sometimes disagreement over Osteosarcoma tumor classification due to cellular heterogeneity in the dataset. Segmentation and classification of histology tissue in H&E stained tumor image datasets is challenging due to intra-class variations and inter-class similarity, crowded context, and noisy data. In recent years, deep learning approaches have led to encouraging results in breast cancer and prostate cancer analysis. In this paper, we propose a Convolutional neural network (CNN) as a tool to improve efficiency and accuracy of Osteosarcoma tumor classification into tumor classes (viable tumor, necrosis) vs non-tumor. The proposed CNN architecture contains five learned layers: three convolutional layers interspersed with max pooling layers for feature extraction and two fully-connected layers with data augmentation strategies to boost performance. We conclude that the use of neural network can assure high accuracy and efficiency in Osteosarcoma classification.

Original language	English (US)
Title of host publication	Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings
Editors	Zhipeng Cai, Ovidiu Daescu, Min Li
Publisher	Springer Verlag
Pages	12-23
Number of pages	12
ISBN (Print)	9783319595740
DOIs	https://doi.org/10.1007/978-3-319-59575-7_2
State	Published - 2017
Event	13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017 - Honolulu, United States Duration: May 29 2017 → Jun 2 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10330 LNBI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017
Country/Territory	United States
City	Honolulu
Period	5/29/17 → 6/2/17

Keywords

Convolutional neural network
Histology image analysis
Osteosarcoma

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-319-59575-7_2

Cite this

Mishra, R., Daescu, O., Leavey, P. J., Rakheja, D., & Sengupta, A. L. (2017). Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network. In Z. Cai, O. Daescu, & M. Li (Eds.), Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings (pp. 12-23). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10330 LNBI). Springer Verlag. https://doi.org/10.1007/978-3-319-59575-7_2

Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network. / Mishra, Rashika; Daescu, Ovidiu; Leavey, Patrick J et al.
Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings. ed. / Zhipeng Cai; Ovidiu Daescu; Min Li. Springer Verlag, 2017. p. 12-23 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10330 LNBI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Mishra, R, Daescu, O, Leavey, PJ , Rakheja, D & Sengupta, AL 2017, Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network. in Z Cai, O Daescu & M Li (eds), Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10330 LNBI, Springer Verlag, pp. 12-23, 13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017, Honolulu, United States, 5/29/17. https://doi.org/10.1007/978-3-319-59575-7_2

Mishra R, Daescu O, Leavey PJ , Rakheja D , Sengupta AL. Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network. In Cai Z, Daescu O, Li M, editors, Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings. Springer Verlag. 2017. p. 12-23. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-59575-7_2

Mishra, Rashika ; Daescu, Ovidiu ; Leavey, Patrick J et al. / Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network. Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings. editor / Zhipeng Cai ; Ovidiu Daescu ; Min Li. Springer Verlag, 2017. pp. 12-23 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{ce61874341a3435792c02f265d32b93d,

title = "Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network",

abstract = "Pathologists often deal with high complexity and sometimes disagreement over Osteosarcoma tumor classification due to cellular heterogeneity in the dataset. Segmentation and classification of histology tissue in H&E stained tumor image datasets is challenging due to intra-class variations and inter-class similarity, crowded context, and noisy data. In recent years, deep learning approaches have led to encouraging results in breast cancer and prostate cancer analysis. In this paper, we propose a Convolutional neural network (CNN) as a tool to improve efficiency and accuracy of Osteosarcoma tumor classification into tumor classes (viable tumor, necrosis) vs non-tumor. The proposed CNN architecture contains five learned layers: three convolutional layers interspersed with max pooling layers for feature extraction and two fully-connected layers with data augmentation strategies to boost performance. We conclude that the use of neural network can assure high accuracy and efficiency in Osteosarcoma classification.",

keywords = "Convolutional neural network, Histology image analysis, Osteosarcoma",

author = "Rashika Mishra and Ovidiu Daescu and Leavey, {Patrick J} and Dinesh Rakheja and Sengupta, {Anita L}",

note = "Funding Information: This research was partially supported by NSF award IIP14 39718 and CPRIT award RP150164. We would like to thank Harish Arunchalam, Bodgan Armaselu and Dr. Riccardo Ziraldo from our group at UT Dallas, and Dr. Lan Ma, University of Maryland, for their helpful discussions. We also would like to thank John-Paul Bach and Sammy Glick from UT Southwestern Medical Center for their help with the datasets. Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017 ; Conference date: 29-05-2017 Through 02-06-2017",

year = "2017",

doi = "10.1007/978-3-319-59575-7_2",

language = "English (US)",

isbn = "9783319595740",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "12--23",

editor = "Zhipeng Cai and Ovidiu Daescu and Min Li",

booktitle = "Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings",

}

TY - GEN

T1 - Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network

AU - Mishra, Rashika

AU - Daescu, Ovidiu

AU - Leavey, Patrick J

AU - Rakheja, Dinesh

AU - Sengupta, Anita L

N1 - Funding Information: This research was partially supported by NSF award IIP14 39718 and CPRIT award RP150164. We would like to thank Harish Arunchalam, Bodgan Armaselu and Dr. Riccardo Ziraldo from our group at UT Dallas, and Dr. Lan Ma, University of Maryland, for their helpful discussions. We also would like to thank John-Paul Bach and Sammy Glick from UT Southwestern Medical Center for their help with the datasets. Publisher Copyright: © Springer International Publishing AG 2017.

PY - 2017

Y1 - 2017

N2 - Pathologists often deal with high complexity and sometimes disagreement over Osteosarcoma tumor classification due to cellular heterogeneity in the dataset. Segmentation and classification of histology tissue in H&E stained tumor image datasets is challenging due to intra-class variations and inter-class similarity, crowded context, and noisy data. In recent years, deep learning approaches have led to encouraging results in breast cancer and prostate cancer analysis. In this paper, we propose a Convolutional neural network (CNN) as a tool to improve efficiency and accuracy of Osteosarcoma tumor classification into tumor classes (viable tumor, necrosis) vs non-tumor. The proposed CNN architecture contains five learned layers: three convolutional layers interspersed with max pooling layers for feature extraction and two fully-connected layers with data augmentation strategies to boost performance. We conclude that the use of neural network can assure high accuracy and efficiency in Osteosarcoma classification.

AB - Pathologists often deal with high complexity and sometimes disagreement over Osteosarcoma tumor classification due to cellular heterogeneity in the dataset. Segmentation and classification of histology tissue in H&E stained tumor image datasets is challenging due to intra-class variations and inter-class similarity, crowded context, and noisy data. In recent years, deep learning approaches have led to encouraging results in breast cancer and prostate cancer analysis. In this paper, we propose a Convolutional neural network (CNN) as a tool to improve efficiency and accuracy of Osteosarcoma tumor classification into tumor classes (viable tumor, necrosis) vs non-tumor. The proposed CNN architecture contains five learned layers: three convolutional layers interspersed with max pooling layers for feature extraction and two fully-connected layers with data augmentation strategies to boost performance. We conclude that the use of neural network can assure high accuracy and efficiency in Osteosarcoma classification.

KW - Convolutional neural network

KW - Histology image analysis

KW - Osteosarcoma

UR - http://www.scopus.com/inward/record.url?scp=85020721422&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020721422&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-59575-7_2

DO - 10.1007/978-3-319-59575-7_2

M3 - Conference contribution

AN - SCOPUS:85020721422

SN - 9783319595740

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 12

EP - 23

BT - Bioinformatics Research and Applications - 13th International Symposium, ISBRA 2017, Proceedings

A2 - Cai, Zhipeng

A2 - Daescu, Ovidiu

A2 - Li, Min

PB - Springer Verlag

T2 - 13th International Symposium on Bioinformatics Research and Applications, ISBRA 2017

Y2 - 29 May 2017 through 2 June 2017

ER -

Histopathological diagnosis for viable and non-viable tumor prediction for osteosarcoma using convolutional neural network

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this